In this project, image processing techniques are used to analyze electron micrographs. To answer important questions in structural biology, it is necessary to obtain relatively high resolution 2- and 3-D structural information about biological macromolecules. Biological specimens can be visualized in the electron microscope using a number of specimen preparation techniques. Cryoelectron microscopy, a relatively new technique, attempts to preserve native structure by surrounding the specimen with a layer of ice. Collaborative studies with LSB, NIAMS are currently under way on a number of projects, whereby electron micrograph images are computationally corrected, combined, averaged, reconstructed, or in some way computationally enhanced to improve the signal-to-noise ratio or to increase the interpretability of the structures being visualized. "Cryo" images are typically lower contrast and require greater computer processing than conventional electron microscopy to achieve satisfactory results. Of particular interest to our research is the understanding of viral structures. At present we are continuing our efforts to investigate the structure of a large animal virus, human herpes simplex virus (type 1). We are completing the localization of the major capsid proteins and attempting to obtain higher resolution structures. Biological material for these herpesvirus reconstructions is provided through a collaboration with researches at the University of Virginia, Charlottesville, and from the Upjohn Co., Kalamazoo. The electron microscopy is performed in LSB, NIAMS. Interpretation of our 3-D reconstructions is performed jointly by all collaborators. A number of other collaborative projects in structural biology are currently in progress. We are using 3-D reconstruction techniques to study the structure of icosahedral L-A virus (from yeast), papillomavirus, and polio virus. We have compared the structures of full (RNA containing) L-A virus with empty L-A virus. In a new study of papillomavirus (in collaboration with NIAMS and NCI), we have verified the known structure of bovine papillomavirus (bpv), and have recently obtained a 3D reconstruction of antibodies to the L1 protein of bpv. We hope to be able to localize the two major proteins of bpv, as well as to understand more of the function and activity of a number of papilloma antibodies.